DESCRIPTION (investigator's abstract): Adolescents, more than other age group, suffer morbidity as a consequence of their behavior. Much research has sought to determine why adolescents take risks, and how they might be influenced to do otherwise. Most theories of risk are psychological in nature, taking the individual as a unit of analysis and hypothesizing individual characteristics as correlates of behavior. One major determinant of behavior overlooked by this approach is that of culture. Consisting of ideas, understanding, beliefs, and subjective knowledge, culture is a determinant of adolescent behavior which does not reside in the individual but is shared by adolescent social groups. The culture of romantic relationships-what adolescents subjectively believe about these relationships-has not been widely studied. As a consequence, interventions targeting adolescents reproductive health often do not contextualize health promotion messages in terms of concepts that adolescents view as important when they think and talk about romantic relationships in their daily lives. In this study, we plan to identify cultural models of romantic relationships based on accounts given by a sample of African American/ Mexican American male/ female adolescents. To identify cultural models, we will elicit cultural cognition in three interview contexts (individual, couple, and peer group) using a Vernacular Term Interview (VTI) procedure. We will then specify the structure and content of cultural models in data from the individual interview using Vernacular Concept Analysis (VCA), in which data is indexed, a subset is extracted, and analytic categories are identified and organized. This method will identify prominent models expressed by respondents of any demographic category or combination of demographic categories. In the second half of the study, we will seek to answer the following comparative questions: 1) what differences are there in cultural models expressed by African Americans and Mexican American adolescents, 2) what differences are there in cultural models expressed my male and female adolescents, and 3) what differences are there in expression and qualification of cultural models by adolescents interviewed in different social contexts (individual, couple, and peer group)? To accomplish this, a uniform coding system reflecting identified cultural models will be applied to data collected in all three interview contexts and statistical procedures will be utilized to answer the three comparative research questions. Findings of this research will have relevance to health promotion efforts related to reducing STD's and unwanted pregnancy.